JP2551787B2 - Variable damping force suspension controller - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a suspension control device that continuously and optimally controls the characteristics of a suspension in a traveling device.

(Prior Art) The inventors of the present invention previously noted that in the suspension of a vibrating body, a slight consumption is required for the purpose of suppressing vibration or providing a vibration-proof effect when vibration is caused by the influence of external force or disturbance. The energy is used to determine the external state such as the road surface or disturbance based on the change in the amount of state accompanying the vibration of the vibrating body, and based on the determination, the optimum gain is selected from the pre-stored gains. The target control force for optimizing the vibration characteristics is calculated by multiplying the gained detected state quantity or physical quantity, and the damping force of the suspension of the vibrating body is controlled to follow the target control force. And a suspension control device for reducing the amount of vibration of the vibrating body have been proposed (see Japanese Patent Application No. 62-108319). This conventional device, as shown in FIG. 2, has a target control force calculation means II.
₁ is equipped with state frequency distribution calculation means II ₁₁ to measure the frequency distribution of state quantities for multiple levels, and based on the measurement results, continuous changes in the external state can be distinguished from the state changes due to sporadic external forces, disturbances, or noise. By making a separate determination, it is possible to avoid erroneously determining a state change caused by a single external force, disturbance, or noise as a continuous state change. Therefore, there is no frequent switching of the target control force due to a single external force, disturbance, or noise, and therefore, the shock accompanying such switching can be reduced, and the vibration acting on the vibrating body can be stably controlled. I have.

(Problems to be Solved by the Invention) However, in the above-described conventional technique, the target control force is calculated and controlled based on the change amount of the state amount of the vibrating body, and the change in the state amount is predicted. Since it is not done, there is a problem that the rise of the control against the sudden change of the state is not quick.

Further, in the related art, since it is not possible to predict the variation of the state quantity assuming the entire movement of the vehicle, there is a problem that the followability to a complicated movement related to each other is low.
Further, the prior art was able to finely respond to the determination of external conditions such as road surface disturbances for each single wheel, but it was not possible to determine the motion state of the entire vehicle from them, and the optimum target control force was determined from the overall motion of the vehicle. There was a problem that I could not get it.

The present invention aims to solve these problems.

That is, one object of the present invention is to provide a suspension control device that can predict the state of a vehicle that should come from the state quantity and the physical quantity and that quickly rises in response to a sudden change in the state.

Another object of the present invention is to provide a suspension control device capable of predicting a change in the state quantity assuming the movement of the entire vehicle and enhancing the followability to a complex movement related to each other.

(Means for Solving Problems) In order to achieve the above object, the present invention detects a physical quantity that affects the characteristics of a suspension supporting a vehicle and shows the movement of the suspension, as shown in FIG. The state detection means I for detecting the state quantity and the state quantity indicating the movement of the vehicle, the detection control force calculation means II ₂ for calculating the detection control force corresponding to the physical quantity detected by the state detection means I, and the state detection means I From the output of the physical movement of the whole vehicle and the physical quantities and external variables such as external forces acting on the suspension and disturbances, the predictive calculation of the upcoming movement of the entire vehicle is performed based on the reference model corresponding to the running movement. , Vehicle condition predicting means II ₁₁ for calculating and outputting an evaluation amount for evaluating the change in the motion condition of the entire vehicle
And a vehicle in which the detected control force is taken into consideration in the reference model so as to control a state quantity having a large change amount to be minimum among changes in the motion state of the vehicle predicted based on the output of the vehicle state prediction means II _11. Gain selecting means II for selecting the optimum feedback control gain calculated based on the control model II
₁₂ and an optimum target control force calculating means for multiplying each gain selected by the gain selecting means II ₁₂ with the physical quantity signal and the state quantity signal which are the outputs of the state detecting means I and adding them.
A control means II having a target control force calculation means II _{1 including} II ₁₃ and a deviation calculation means II ₃ for calculating a deviation between the target control force and the detected control force, and an output of the control means II. Equivalent to the driving means III for power amplifying the deviation signal of the control force, and the control force corresponding to the deviation of the actual detected control force from the target control force considering the external force or disturbance acting on the suspension based on the power-amplified output. It is composed of actuator means IV that continuously and variably controls the suspension characteristics so that it can accurately judge the continuous change in the vehicle motion state from the degree of change in the state quantity or physical quantity of the suspension of the entire vehicle and each wheel. Then, by calculating the optimum target control force using the gain corresponding to it, the optimum target control force that is not affected by the sudden external force or the disturbance component with respect to the vehicle motion state can be obtained. It is a damping force variable suspension control device that generates and continuously and optimally controls suspension characteristics.

(Operation) The operation and effect of the present invention having the above configuration are as follows. That is, first, from the physical quantity Z and the state quantity X representing the external state such as the external force acting on the suspension and the movement of the entire vehicle which is the output of the state detection means I, and the disturbance,
An evaluation amount for predicting and calculating the coming motion state of the entire vehicle based on a reference model corresponding to the motion state during traveling, and evaluating the change in the coming motion state of the entire vehicle, that is,
The control gain is calculated by the evaluation functions J _{1 to} J ₄ described later and output.

The gain selecting means II ₁₂ detects and controls the reference model so as to minimize the state quantity having a large change amount in the change of the motion state of the vehicle predicted based on the output of the vehicle state predicting means II _11. The optimum feedback control gain calculated based on the vehicle control model considering the force is selected.

Optimal target control force calculation means using selected control gain
II ₁₃ calculates the optimum target control force, generates the optimum target control force that is not affected by components such as the motion state of the vehicle and single external forces and disturbances, and outputs its output with drive means III. The actuator means IV attached to the suspension is driven to continuously and optimally control the suspension characteristics.

Next, the principle of calculating the optimum feedback gain in the vehicle state predicting means II ₁₁ of the present invention will be described in detail.

The motion of the vehicle can be expressed by the following motion equations for the three motions of the lateral translation motion, the yaw motion, and the rolling motion.

ΣI _Z1 = ΣN _ψ (2) I _x = 2 (M _uf Z _f + M _ur Z _r ) (v + V ₀ γ−gφ) + ΣN _φ
(3) where ΣM: gross vehicle weight, M _uf , M _ur : front and rear unsprung mass, V ₀ : vehicle speed, γ: yaw angle, φ: roll angle, F _si :
Side force, I _z : yaw inertial ratio, I _x : roll inertial ratio, N _ψ : yaw moment, N _φ : roll moment,
g: gravitational acceleration, v: lateral translation speed.

Further, when the motion of the vehicle during acceleration / deceleration is assumed, a 4-degree-of-freedom model as shown in FIG. 3 is considered as follows. Considering a three-mass system four-degree-of-freedom control system of sprung mass and unsprung mass, the following equations that give a reference model hold. In the figure, m ₁ , m ₂ , m ₃ , k _T1 , k _T2 , k ₁ , k ₂ , c ₁ , c ₂ are the first
These are the quantities shown in the table, where Z ₁ and Z ₂ are the unsprung displacements,
Z ₁₀ and Z ₂₀ are displacements of the road surface, Z _3f and Z _3r are sprung displacements, and θ is a steering angle.

m _{1 1} = -k _T1 (Z ₁ -Z ₁₀ ) + c ₁ ( _{3f -1} ) + k ₁ (Z _3f -Z ₁ ) (4) m _{2 2} = -k _T2 (Z ₂ -Z ₂₀ ) + c ₂ ( _{3r - 2) + k 2 (} Z 3r -Z 2) (5) m 3 3 = -c 1 (3f - 1) -k 1 (Z 3f -Z 1) -c 2 (3r - 2) -k 2 ( Z _3r −Z ₂ ) (6) I = l _f {c ₁ ( _3f − ₁ ) + k ₁ (Z _3f −Z ₁ )} −l _r {c ₂ ( _3r − ₂ ) + k ₂ (Z _3r −Z ₂ )} (7) Z _3f = Z ₃ + (-l _f ) θ, Z _3r = Z ₃ + l _r θ _3f = ₃ + (-l _f ), _3r = ₃ + l _r (8) Where, vehicle speed, steering wheel Angle, front road surface displacement, and vehicle height displacement (relative displacement) are V ₀ , δ _f , Z ₁₀ , Z ₂₀ , Z _3f −Z ₁ , Z _3r −
Assume a predicted motion that is a reference of the vehicle in the case of Z ₂ .

The predicted motion in the reference model is here, And can be arranged as the following equation. (Vehicle speed V ₀ , steering wheel angle δ _f as input) Therefore, the predicted motion evaluation amount can be estimated and calculated from the equations (4) to (7) and the equations (13) to (15).

 This calculation means is as follows.

 Expressions (13) to (15) are expressed as the following expressions.

For example, the predicted roll angle φ _∞ is Then, applying Cramer's formula, it becomes the following formula.

φ _∞ = D _φ / D (19) Similarly, the predicted lateral translational velocity v _∞ and the predicted yaw angle γ _∞
Is required.

Below is a summary of other predictive exercise evaluation quantities,
It looks like Table 1.

That is, the instantaneous value of the steering wheel δ _f , which changes momentarily under the vehicle speed V ₀ , and the predicted motion evaluation amount corresponding to the front road surface displacement are estimated and calculated.

Next, the principle of calculating the optimum control gain based on the predicted evaluation amount will be described.

Here, assuming that the target control forces are u ₁ and u ₂ and the actual operating forces are f ₁ and f ₂ , since there is usually an operation delay in the relationship between them, the following equation holds.

Further, the following y ₁ and y ₂ are introduced to organize the equations (4) to (7).

Front wheel relative displacement y ₁ ＝ Z ₁ −Z _3f ＝ Z ₁ −Z ₃ ＋ l _f θ Rear wheel relative displacement y ₂ ＝ Z ₂ −Z _3r ＝ Z ₂ −Z ₃ ＋ l _r θ Z ₁ ＝ Z _3f ＋ y ₁ ＝ Z ₃ + y ₁ -l _f θ, Z ₂ = Z ₃ + y ₂ + l _r θ When re-arranged and summarized, the following formula is established.

Perform variable conversion.

(X ₁ , x ₂ , x ₃ , x ₄ , x ₅ , x ₆ , x ₇ , x ₈ , x ₉ , x ₁₀ ) = (y ₁ , ₁ , y ₂ , ₂ , Z ₃ , ₃ , θ, , f ₁ , f ₂ ) (21) Equations (1 ″) to (6 ″) are represented by the following state equations. ₁ = x _{2 3} = x _{4 5} = x _{6 7} = x ₈ However, here, The equation of state is as follows.

Here, the A and B matrices are as follows:

In the above 4-degree-of-freedom model, assuming the specifications shown in Table 2 (a), the equation of state describing the pitch bounce motion during acceleration / deceleration of the vehicle is given,
Assuming the specifications shown in the column (b) of the table, a state equation describing the roll bounce motion when steering the vehicle or when overcoming a rudder in left and right phases is given.

When obtaining the optimum state feedback gain, the following state evaluation function J _i is assumed in consideration of various vehicle state changes.

First, when the ride comfort during steady running, that is, the vibration level on the spring is lowered, and when the unsprung vibration level is lowered in consideration of the grounding property when running on a rough road surface, further, the roll during steering It is considered that the sprung unsprung relative displacement is reduced in consideration of the change of motion and the roll rigidity distribution is optimized by further distributing the amount of change in the change of front and rear.

In addition, the sprung mass m of the vehicle model in this model is
₃ needs to be changed according to the loaded weight, and the spring constants k ₁ and k _{2 of the} suspension also change depending on the loaded weight when a gas-liquid fluid spring is set as the suspension spring, so the specifications in the model should be set promptly. There is a need.

After setting the above-mentioned specifications to the vehicle state, the following evaluation function J _i is set according to the predicted motion state of each vehicle.

Predicted motion evaluation amount when assuming riding comfort during steady running
_The evaluation function J _i is set as follows according to _∞ .

Here, ρ ₁ , ρ ₂ and q ₁ , q ₂ are weighting factors for multiplying the control amount and the evaluation amount, respectively.

When considering the ground contact property at the same time, the predicted motion evaluation value
_The evaluation function is as follows according to _1∞ and _2∞ .

Further, when considering the changeover during steering, according to the predicted motion evaluation quantities φ _∞ , γ _∞ , v _∞ , Furthermore, when considering the distribution of front and rear, according to the predicted motion evaluation amount _∞ , And so on.

 Generally, the evaluation function J is expressed by the following equation.

That is, based on the state equation shown in equation (8),
The optimal feedback control system is constructed by solving the optimal regulator under the evaluation function as shown above.

Here, the optimal control force that minimizes the evaluation function J Is its state quantity It is expressed by the following equation as a function of.

However, As is well known by the optimal control theory, state feedback gain Is given by

Here, P is a positive definite solution of the following Riccati matrix equation.

^{PA + A T P-PBR -1} B T P + Q = 0 (31) calculations proceed in the form of finding the steady-state solution obtained by solving the following equation in the initial condition P (0) = 0.

= The ^{-PA-A T P + PBR -1} B T P-Q (32) over the deployment, the optimal feedback gains K ₁ ~K
₁₀ is obtained.

 The above results are generally shown as follows.

The time-series optimum target control force u considering the vibration of the vibrating body generated by the external force or the disturbance acting on the suspension supporting the vibrating body m is, for example, the motion equation in FIG. Is as follows.

m = u (x ,,) (33) where x is suspension displacement due to external force or disturbance, is suspension speed, and is acceleration given to the vibrating body. That is, the target control force u is a function of x ,.

Here, if the target control force u is further shown in a general form,
It looks like this:

Here, g _i is a contribution gain coefficient for giving optimum vibration suppression, and x _i is all state quantities that can describe the present vibration system. In general, this includes transmission between the suspension components. That is, the optimum target control force u forms a so-called instantaneous state feedback control system by instantaneously and instantaneously detecting the state physical quantity x _i of the vibrating body m and giving the coefficient g _i according to the respective contributions. It is possible to give optimum vibration suppression to the present mass vibration system.

With respect to this target control force u, a physical quantity f acting on the suspension is detected by a sensor, and its deviation ε (= u-
The output of f) is power-amplified by the drive means III, the actuator means IV attached to the suspension is driven, and the physical quantity f is continuously controlled.

According to the present invention, the vehicle state predicting means II ₁₁ of the target control force calculating means II ₁ is based on the signal of the state quantity or the physical quantity of the suspension based on the reference model representing the standard state of the entire vehicle. Since the prediction calculation of the vehicle state is performed, the rise of the control for the sudden change of the state becomes fast, and the followability for the complicated movements related to each other becomes high. Therefore, the optimum target control force can be set from the entire motion of the vehicle, control that is in accordance with the motion state of the vehicle can be performed, and vibration reduction of the vehicle can be controlled stably.

[Embodiment]

FIG. 4 is a block diagram showing an embodiment of the present invention. In the basic configuration of the present invention shown in FIG. 1, the vehicle state predicting means II ₁₁ is the output of the state detecting means I and the movement of the entire vehicle. Based on a reference model corresponding to an average motion state during running, which corresponds to a motion state determined by a physical quantity Z and a state quantity X that represents an external state such as an external force and a disturbance acting on the suspension, the movement state of the entire vehicle to come Based on the output of the predictive state calculating means II ₁₁₁ for predictive calculation and the output of the predictive state calculating means II ₁₁₁ , the movement of the vehicle accompanied by a change in at least two state quantities in which the change in the moving state of the vehicle is predicted to be the largest. Predictive state determination means for determining the state
II ₁₁₂ , and the gain means II ₁₂ controls the reference model in consideration of the detected control force so as to control the variation of the motion state of the vehicle determined by the prediction state determination means II ₁₁₂ to the minimum. Is configured to select the optimum feedback control gain calculated based on

The operation and effect of the embodiment of the present invention having such a structural feature will be described.

The predictive state calculating means II ₁₁₁ is a physical quantity Z and a state quantity X which represent the output of the state detecting means I and represent an external state such as an external force and a disturbance acting on the motion and suspension of the entire vehicle.
Based on the reference model corresponding to the average motion state at the time of traveling corresponding to the motion state determined from the above, the motion state of the entire coming vehicle is predicted and calculated.

The predicted state determination means II ₁₁₂ is based on the output of the predicted state calculation means II ₁₁₁ , and is based on the output of the predicted state calculation means II _111. Left and right different phase fluctuation state, steering state,
The acceleration / deceleration state, the steady running state, etc. are determined.

The gain means II ₁₂ is an optimum value calculated based on the vehicle control model in which the detected control force is taken into consideration in the reference model, in order to control so as to minimize the fluctuation of the motion state of the vehicle determined by the prediction state determination means II _112. Select the feedback control gain.

According to this embodiment, the predicted state calculation means II ₁₁₁ performs the state prediction calculation of the entire vehicle based on the signals from the various sensors installed in the vehicle, so that quick control is possible and the predicted state determination means With II ₁₁₂ , the motion state of the entire vehicle can be discriminated based on the reference model, so that precise optimal control of vibration is possible.

(Example) FIG. 5 (a), (b) and FIG. 6 are schematic configuration diagrams of an example belonging to the embodiment of FIG. 4 of the present invention.
5 (a) and 5 (b) show parts corresponding to the driving means III and the actuator means IV in the configuration of FIG. FIG. 6 shows an overall block diagram of this embodiment.

In these figures, 1 indicates a reserve tank that stores oil as a working fluid, and 2fr, 2fl, 2rr, 2r
Reference characters l denote actuators provided for the right front wheel, left front wheel, right rear wheel, and left rear wheel, respectively, which are not shown in the figure. As shown in FIG. 5 (a), each actuator is composed of a cylinder 3 and a piston 4 which are respectively connected to the vehicle body and suspension arm of the vehicle, and a cylinder chamber as a working fluid chamber defined by these. By supplying / discharging the oil to / from 5, the vehicle height at each corresponding position can be increased or decreased.
It should be noted that the actuator increases or decreases the vehicle height at a corresponding position by supplying or discharging a working fluid such as oil to or from the working fluid chamber, and the pressure in the working fluid chamber also increases or decreases in accordance with the bouncing or rebound of the wheel. Any device such as a hydraulic ram device may be used as long as it is configured as described above.

The reserve tank 1 is connected to a branch point 11 by a conduit 10 having an oil pump 6, a flow control valve 7, an unload valve 8 and a check valve 9 along the way. The pump 6 is driven by the engine 12 to pump oil from the reserve tank 1 and discharge high-pressure oil, and the flow control valve 7 controls the amount of oil flowing in the conduit 10 downstream thereof. It is designed to control the flow rate. The unload valve 8 detects the pressure in the conduit 10 downstream of the check valve 9, and when the pressure exceeds a predetermined value, returns the oil to the conduit 10 upstream of the pump 6 via the conduit 13. As a result, the pressure of the oil in the conduit 10 on the downstream side of the check valve 9 is maintained below a predetermined value. Check valve 9 is a branch point
The oil is prevented from flowing backward in the conduit 10 from 11 toward the unload valve 8.

The branch point 11 is connected to the cylinder chamber 5 of the actuators 2fr and 2fl by conduits 20 and 21 each having check valves 14 and 15, solenoid on-off valves 16 and 17, and electromagnetic flow control valves 18 and 19 on the way. There is. Further, the branch point 11 is connected to the branch point 23 by a conduit 22, and the branch point 23 is provided with check valves 24 and 25, electromagnetic switching valves 26 and 27, and
The conduits 30 and 31 having the electromagnetic flow control valves 28 and 29 are connected to the cylinder chambers 5 of the actuators 2rr and 2rl, respectively.

Thus, the cylinder chamber 5 of the actuators 2fr, 2fl, 2rr, 2rl is selectively supplied with oil from the reserve tank 1 via the conduits 10, 20 to 22, 30, 31 and in that case. The oil supply and the flow rate thereof are appropriately controlled by controlling the on-off valves 16, 17, 26, 27 and the flow rate control valves 18, 19, 28, 29, respectively.

The portions of the conduits 20 and 21 between the flow control valves 18 and 19 and the actuators 2fr and 2fl are electromagnetic flow control valves 32 and 33, electromagnetic switching valves 34 and 35, respectively, on the way.
By means of the conduits 36 and 37 having the above, a return conduit 38 communicating with the reserve tank 1 is connected. Similarly, the conduit
The portions between 30 and 31 between the flow control valves 28 and 29 and the actuators 2rr and 2rl are connected to the return conduits by conduits 43 and 44 having electromagnetic flow control valves 39 and 40 and electromagnetic opening / closing valves 41 and 42, respectively. It is connected to 38.

Thus, the oil in the cylinder 5 of the actuators 2fr, 2fl, 2rr, 2rl is selectively discharged to the reserve tank 1 via the conduits 36 to 38, 43, 44, and the oil discharge in that case. And its flow rate, as will be described later in detail, on-off valves 34, 35, 41, 42 and flow control valve 3 respectively.
It is appropriately controlled by controlling 2,33,39,40.

In the illustrated embodiment, the on-off valves 16, 17, 26, 27, 34, 35,
41 and 42 are normally-closed type on-off valves.When the corresponding solenoids are not energized, the closed state is maintained as shown in the figure to disconnect the corresponding conduits and energize the corresponding solenoids. When the operation is performed, the valve is opened to allow the communication of the corresponding conduit. Further, the flow control valves 18, 19, 28, 29, 32, 33, 39, 40 change the throttle degree by changing the voltage of the drive current or the duty of the current supplied to the corresponding solenoids, As a result, the flow rate of oil flowing through the corresponding conduit is controlled.

Accumulators 45 to 48 are connected to the conduits 20, 21, 30, and 31 at positions upstream of the check valves 14, 15, 24, and 25, respectively. Each accumulator is composed of an oil chamber 49 and an air chamber 50 which are separated from each other by a diaphragm, and compensates for the pulsation of the oil by the pump 6 and the pressure change in the conduit 10 due to the action of the unload valve 8 and responds. Conduit 2
It is designed to have a pressure accumulating action on the oil in 0, 21, 30, 31.

In the portions of the conduits 20, 21, 30, 31 between the flow control valves 18, 19, 28, 29 and the corresponding actuators, the main springs are provided by conduits 55-58 having variable throttle devices 51-54 in the middle. 59 to 62 are connected, and conduits 67 to 70 having normally open type opening / closing valves 63 to 66 are provided in the middle of the conduits 55 to 58 between the variable throttle device and the main spring, respectively. Secondary springs 71 to 74 are connected. The main springs 59 to 62 each include an oil chamber 75 and an air chamber 76 which are separated from each other by a diaphragm, and similarly, the sub springs 71 to 74 respectively include an oil chamber 77 and an air chamber 78 which are separated from each other by a diaphragm. Is made up of.

Thus, as shown in FIG. 5 (a), when the volume of the cylinder chamber 5 of each actuator changes due to the bouncing and rebound of the wheels, the cylinder chamber and the oil chamber
Oils in 75 and 77 mutually flow through the variable expansion devices 51 to 54, and the vibration resistance is exerted by the flow resistance at that time. In this case, the damping degree C of each variable throttle device is controlled by the corresponding motors 79 to 82, so that the damping force C can be continuously and steplessly switched, and the on-off valves 63 to 66 can be switched. Corresponding motor 83
By selectively opening and closing by ~ 86, the spring constant k can be switched between high and low. The motors 79 to 82 and the motors 83 to 86 use a vehicle speed sensor 95, a steering angle sensor 96, a throttle opening sensor 97, and a braking sensor 98 to reduce the nose dive, scooter, and roll of the vehicle, as described later. It is controlled.

Further, vehicle height sensors 87 to 90 are provided at positions corresponding to the actuators 2fr, 2fl, 2rr, 2rl, respectively. These vehicle height sensors detect the vehicle height at a corresponding position by measuring the relative displacement between the cylinder 3 and the piston 4 or a suspension arm not shown in the figure, and indicate the vehicle height. The signal is output to the electronic control unit 102 shown in FIG.

An electronic control unit 102 corresponding to the control means II in the embodiment of FIG. 4 includes a microcomputer 103 as shown in FIG. Microcomputer 1
Reference numeral 03 may have a general configuration as shown in FIG. 6, which includes a central processing unit (CPU) 104, a read only memory (ROM) 105, and a random access memory (RA).
M) 106, input port device 107 and output port device 108
And are connected to each other by a bi-directional common bus 109.

The input port device 107 indicates whether the vehicle height selected by the vehicle height selection switch 110 provided in the vehicle compartment and operated by the driver is high (H), normal (N), or low (L). The switch function signal shown is input. Further, the input port device 107 has a vehicle height sensor 8
Actual vehicle height Hfr, Hf detected by 7,88,89,90 respectively
l, Hrr, Hrl signal, vehicle speed sensor 95, steering angle sensor 96,
Signals indicating vehicle speed V, steering angle δ (right turn is positive), throttle opening θ, and braking state detected by throttle opening sensor 97 and braking sensor 98 respectively correspond to amplifiers 87a-90a, 95a- It is designed to be input via the 98a, the multiplexer 111, and the A / D converter 112.

The ROM 105 stores fixed data necessary for arithmetic processing by the CPU 104, programs for executing the functions of the control means, and the like. The data includes, for example, a map for performing a vehicle state quantity prediction calculation, a constant for determining a prediction state, a set value for gain selection, and the like. The reference vehicle heights Hhf and Hhr, Hnf and Hnr, Hlf and Hlr (Hhf as the target vehicle heights of the front wheels and the rear wheels when the vehicle height selection switch 110 is set to high, normal or low.
>Hnf> Hlf, Hhr>Hnr> Hlr).

The RAM 106 includes various sensors 87-
It stores various data necessary for arithmetic processing of the CPU 104 such as detection signals from 90 and 94 to 98, arithmetic results and the like.

The CPU 104 performs calculation processing such as prediction state calculation, prediction state determination, gain selection, optimum target control force calculation, detection control force calculation, deviation calculation, etc. according to the program stored in the ROM 105, and based on the calculation result, each actuator To the on-off valve and the flow control valve provided corresponding to the output port device 108 and the corresponding D / A converters 117a-1
17h and 118a to 118h, amplifiers 119a to 119h and 120a to 12
The output port device 108 corresponds to the motors 79 to 82 that selectively output the control signal via 0h and that drive the variable throttle devices 51 to 54 and the motors 83 to 86 that drive the on-off valves 63 to 66, respectively. A control signal is selectively output via the D / A converters 121a to 121h and 123a to 123h and the amplifiers 122a to 122h and 124a to 124h.

On the display 116 connected to the output port device 108, the reference vehicle height selected by the vehicle height selection switch 110 is high,
Whether it is normal or low is displayed.

Next, referring to the flowchart shown in FIG. 7,
The operation of the apparatus of the embodiment of the present invention shown in FIGS. 5 (b) and 6 will be described.

After the initial setting (step P ₁ ), the signals from the respective sensors shown in FIG. 6 are read (step P ₂ ).

Then performed vehicle state quantity prediction calculation (Step P _3). Here, an arithmetic expression for calculating the change prediction of the vehicle state quantity is provided by performing signal processing from the related sensors for acceleration, steering, protrusion step and deceleration, which will be described later.

Next, the calculation of the vehicle state quantity (step P ₄ ) is performed.
That is, the vehicle state quantity is calculated by a predetermined calculation by the state quantity sensor described later attached to the suspension system for each of the four wheels of the vehicle, which changes moment by moment.

Then, it is judged running condition of the vehicle (Step P _5).
Here, it is determined whether the vehicle is in the running state,
If it is less than a predetermined vehicle speed, it initializes the damping valve position (step P _6).

On the other hand, after the judgment is running state, it determines left and right pressure of the different phase variation or the steering state or the vehicle, or deceleration state (step P ₇ ~P _9).

That is, in the out-of-phase variation determination (step P ₇ ), the detection signal of the front road surface sensor 94 is compared with a predetermined value, and when it becomes equal to or more than the predetermined value, it is determined that the swell road surface state is entered.

In the steering determination (step P ₈ ), the predicted roll angle is obtained from the three kinds of signals by the differential value of the signal of the vehicle speed sensor 95 and the signals of the steering angle sensor 96 and the steering sensors 96, and the value is set as a predetermined value. By comparison, it is determined that the steering turning state is entered when the value exceeds a predetermined value.

Acceleration judgment (step P ₉ ) is based on the throttle opening sensor 97
Signal or a value obtained by differentiating the sensor signal is determined, and when the value becomes equal to or more than a predetermined value, it is determined that the acceleration state is entered. Further, the deceleration judgment (Step P ₉₎ obtains a braking sensor 98 or a value obtained by differentiating the sensor signal, it is assumed that the value enters the deceleration state when equal to or above a prescribed value. When neither acceleration nor deceleration has occurred, it is determined that the vehicle is traveling at a constant speed.

Then, based on each judgment, the optimum gain for calculating the target control force is calculated according to the predicted state (step
P ₁₀ ~P _13). Here, assuming an optimal model in advance, in steps P ₁₀ and P ₁₁ , roll bounce control, and in steps P ₁₂ and P ₁₁ ,
_{In step 13} , the optimum feedback gain for pitch / bounce control is calculated.

Then, in accordance with any of the determination in Step P ₁₀ to P _13, the optimum feedback as a result of the state quantity prediction calculation
K _{1 to} K ₁₀ are output, and the optimum gain of the control target force for each wheel is calculated (step P ₁₄ ).

Then, the respective gains determined in step P _14, the two front wheels and two rear wheels, or left two-wheel and right 2 wheels separately one set, and outputs as a valve drive signal of an actuator of the two-wheel.

The principle of how the single-wheel portion of the control device of the present invention is controlled by the drive signal obtained as described above will be described. In the embodiment of FIG. 6, the control means II is all executed by the microcomputer 103 including the CPU 104, ROM 105, RAM 106, etc.
In the figure, for convenience of explanation, only the vehicle state predicting means II ₁₁ and the gain selecting section II ₁₂ are shown to be processed by the microcomputer 270.

As shown in FIG. 8, the state detecting means I includes a suspension arm 262 and a vehicle body frame 2 for rotatably supporting the wheels of the suspension for one of the four wheels.
A potentiometer 210 that is inserted between the potentiometer 210 and 63 to detect a relative displacement, an amplifier 220 that is connected to the potentiometer 210 and outputs a signal representing a relative displacement y ₁ between an axle and a vehicle body when the automobile is running, and an amplifier 220. Differentiator 230 that differentiates the relative displacement y ₁ that is output to detect relative speed ₁ , a pressure sensor 211a that is attached to the hydraulic cylinder 310 to detect the wheel load acting, and a wheel load w is detected from the pressure. Amplifier 221a, a pressure sensor 211b for detecting the damping force by being attached to the inlet of the oil chamber of the accumulator 320, and an amplifier connected to the pressure sensor 211b and amplifying its output.
221b, a differential amplifier 231 that detects the damping force f ₁ as the difference between the output of the amplifier 221b and the output of the amplifier 221a, an acceleration sensor 212 that is mounted on the vehicle body to detect acceleration, and is connected to the acceleration sensor 212. An amplifier 222 that amplifies, an integrator 232a that integrates the output to detect the sprung mass velocity ₃ , and an integrator 23 that further integrates the output to detect the sprung displacement Z _3.
2b, a temperature sensor 213 that is installed in the gas chamber of the accumulator 320 to detect the gas temperature t, an amplifier 223 that is connected to the temperature sensor and amplifies the sensor output, and a vehicle speed V ₀ that is installed on the output shaft of the automobile mission. It comprises the vehicle speed sensor 214 for detecting, a displacement sensor 256, and an amplifier 224 for outputting a signal representing the displacement. The displacement sensor 256
Linear actuator 255 as shown in FIG. 10 (b)
In the actuator means IV including the valve body 259, the displacement of the spool 258, which serves as a variable orifice, is detected by continuously opening and closing the oil passage 150.

The vehicle state prediction means II ₁₁ and the gain selection means II ₁₂ are
An input unit 271 that takes in the vehicle speed V ₀ , relative displacement y ₁ , wheel load w, and gas temperature t, and an arithmetic processing unit that predicts a change in the state quantity and determines the state based on the inputs and selects an optimum gain. 272, a storage unit 273 that stores the optimum gain, each calculation method of the calculation processing unit 272, and constants necessary for calculation in advance.
And a microcomputer 270 including an output unit 274 that outputs the optimum gain selected by the arithmetic processing unit 272. In the embodiment shown in FIG. 6, the optimum target control force calculating means II ₁₃ , the deviation calculating means II ₃ , the sign adjusting means II ₄ , and other calculating parts are also configured by the function of the microcomputer 270.

Microcomputer 27 corresponding to one of the four wheels
The function performed by 0 will be described in detail with reference to the flowchart of FIG.

To capture the output of the vehicle sensor 95 (step P ₂ ) and detect the weight of the vehicle body that changes depending on the number of passengers or the loaded luggage, that is, the sprung mass of the gas-liquid suspension, that is, the sprung mass m of the gas-liquid suspension. Then, the signal from the wheel load sensor is processed, and the spring constants k ₁ and k ₂ at the load change are calculated.

Replacing the wheel suspension with a linear four-degree-of-freedom model in advance, assuming the active control suspension, and using the linear square type optimal control method, the relative displacement y, the relative velocity, and the sprung displacement are combined with k ₂ and m. Z ₃ , sprung speed
₃ , the optimum gains G _{1 to} G ₁₀ for the damping force f ₁ are calculated and read from the stored storage unit 273 and output from the output unit 274.

That is, the relative displacement y ₁ acquired in step P ₂₀ is differentiated to calculate the relative velocity ₁ (step P ₂₃ ). Further, the sprung speed ₃ is calculated from the acceleration ₃ fetched in step P ₂₁ through an integration filter (step P ₂₄ ). Furthermore, (step P ₂₅₎ through a second integrating filter from sprung speed obtained in step P _24, and calculates the sprung displacement Z _3. Next, the coefficient K (x _s ), n (x _s ) is obtained from the map of steps P ₂₆ and P ₂₇ from the valve stroke meter output x _s obtained in step P ₂₂ , and the control force is calculated from these values ( Step P ₂₈ ).

The target control force is calculated from the above signals and the gain map (step P ₁₄ ) (step P ₂₉ ).

The optimum target control force calculation means II ₁₃ is adapted to output the optimum gains G _{1 to} G output from the output section 274 of the microcomputer 270.
_{From 10} and the corresponding status signal, 10 multipliers 241 for each wheel for calculating the optimum target control force u are calculated according to the following equation.
It consists of ~ 245 and an adder 250. Note that FIG. 8 shows only five multipliers.

_{u = G 1 · y 1 +} G 2 · 1 + G 3 · z 3 + G 4 · 3 + G 5 · f 1 + G 6 · y 2 + G 7 · 2 + G 8 · θ + G 9 · + G 10 · f 2 (35) deviation computing The means II ₃ comprises a deviation device 251 which calculates a deviation ε from the damping force fc to be controlled with respect to the optimum target control force u output from the optimum target control force calculating means II ₁₃ .
This corresponds to Step P _30.

The sign adjusting means II ₄ comprises a multiplier 252 which multiplies the output ε of the deviation device 251 by the suspension relative speed. When performing the damping force control according to the deviation ε with respect to the target control force u, the multiplier 252 determines whether or not the deviation ε with respect to the target control force can be controlled by the damping force, and when the control is possible, Outputs a signal that determines the direction of increase / decrease in damping force.
When the control is impossible, the damping force is reduced and a signal in the direction of approaching zero is output. This is step P
Equivalent to ₃₁ .

The code adjusting function of the multiplier 252 will be described with reference to Table 2 and FIG. When the target control force u is positive in the vertical direction with respect to the vehicle body and the relative speed of the suspension is positive in the contraction direction of the gas-liquid suspension, both the target control force u and the relative speed are in the same direction, For example, when the piston of the hydraulic cylinder 310 moves upward (forward direction) and the target control force u is also upward (forward direction),
Since the oil in the hydraulic cylinder 310 flows into the accumulator 320 through the orifice 330 in proportion to the relative speed, the pressure in the hydraulic cylinder 310, that is, the damping coefficient is increased by changing the opening degree of the orifice 330 by the control signal. The magnitude of the damping force fc (in the positive direction) can be changed. In this case, when the output ε of the deviation device 251 is positive (u> fc), the orifice opening is closed, the damping coefficient is increased to increase the damping force, and when ε is negative (u <fc), it is opened. Then, a control signal for reducing the damping force by reducing the damping coefficient may be output. When the piston of the hydraulic cylinder 310 moves downward (negative direction) and the target control force u is also downward (negative direction), oil flows from the accumulator 320 through the orifice 330, conversely to the above. Since it flows into the inside, the magnitude of the downward (negative direction) damping force fc can be changed by similarly controlling the orifice opening. Also in this case, ε is positive (-u>
At −fc), the orifice opening is set to the opening direction, the damping coefficient is reduced to reduce the damping force, and when ε is negative (−u <fc), it is set to the closing direction and the damping coefficient is increased to make it equivalent to the suspension. It suffices to output a control signal that increases the damping force acting on. Therefore, when the target control force u and the suspension relative speed are in the same direction, the damping force fc can be controlled based on the target control force u. On the other hand, when the relative speed of the target control force u is opposite, for example, the piston of the hydraulic cylinder 310 moves upward (positive direction) and the target control force u is downward (negative direction), the oil of the hydraulic cylinder 310 is Since it flows into the accumulator 320 via the orifice 330, if the orifice opening is set to a certain opening (no control), an upward (forward) damping force acts together with the relative speed, and the target The damping force cannot be controlled based on the control force u.

Therefore, if the orifice opening is fully opened by a control signal and the damping coefficient is minimized to reduce the damping force fc in the positive direction that acts equivalently on the suspension, the target control with respect to the damping force fc when no control is performed. This is equivalent to applying a force in the direction of force u and reducing it. The output ε (= u-fc) of the deviation device 251 at this time is always positive because the target control force u is negative and fc is positive because it is in the same direction as the relative speed.

Further, even when the piston of the hydraulic cylinder 310 moves downward (negative direction) and the direction of the target control force is upward (positive direction), the damping force is controlled based on the target control output u in the same manner as above. Therefore, it is desirable that the orifice opening be fully opened by the control signal to minimize the damping coefficient to reduce the damping force equivalently acting on the suspension. Output ε (= u-fc) of the deviation device 251 at this time
Is negative because the target control force u is positive and fc is in the same direction as the relative speed, and therefore is always positive. Therefore, when the target control force u and the relative speed are in the opposite directions, the damping force cannot be controlled based on the target control force u. Therefore, the orifice opening can be fully opened by the control signal to reduce the damping force. It will be good.

As described above, Table 3 summarizes the control directions of the damping force and the orifice opening with respect to the output ε of the deviation device 251 in each state. In order to basically achieve this logic, the sign of ε is multiplied by the sign of the suspension relative speed in the same direction as the damping force, and the output becomes a control signal corresponding to the control direction of the orifice opening. . Here, it suffices that the control signal determines the increasing / decreasing direction of the damping force, and the deviation ε with respect to the target controlling force
In order to improve the ratio of noise to signal, that is, the SN ratio, ε obtained by multiplying ε directly by the relative speed in the multiplier 252.
Was used as the control signal.

The integrating means II ₅ is composed of an operational amplifier, an integrator 253 composed of a resistor R and a capacitor C that determine an integral gain, and integrates the output ε of the multiplier 252 with time to obtain a target control force u.
Offset of the deviation ε from the damping force fc against (Residual deviation)
To eliminate this, the damping force of the suspension is detected and fed back as an integral input, and from the viewpoint of control system responsiveness and stability, the integral gain K _K (= 1 / CR)
Was set to K _K = 2400. Moreover, in order to prevent the drift of the integrator itself, its output was fed back to the input by a resistor.

The driving means III comprises a driving circuit 254 which negatively feeds back the spool displacement signal of the actuator means IV to the output of the integrator 253 and outputs a current proportional to the deviation signal.

As shown in FIG. 10B, the actuator means IV includes a valve body 259 integrally formed with a suspension arm 262 and a hydraulic cylinder 310 of a gas-liquid fluid suspension attached to a vehicle body frame 263, and an oil chamber of an accumulator 320. An oil passage 150 that communicates with the oil chamber of the hydraulic cylinder 310 through the valve body 259, a spool 258 that continuously opens and closes the oil passage 150 to form a variable orifice, and a linear actuator 255 integrated with the spool 258. Of the moving coil 257, a permanent magnet 260 that gives a force that acts on the moving coil 257 according to the current that is the output of the drive circuit 254 that flows through the moving coil 257, and a force that acts on the moving coil by attaching to the linear actuator 255. A displacement sensor 256 for detecting the displacement of the spool 258, and an amplifier 224 for outputting a signal representing the displacement.

The movement of the spool 258 with respect to ∫εdt obtained by time integration of the output ε of the multiplier 252 of the control means II, which is the control input of the actuator means, will be described with reference to FIG. First
The control input ∫εdt is plotted on the horizontal axis of Fig. 1, and the spool displacement x is plotted on the vertical axis.
_The damping coefficient C with respect to _s , the orifice opening a, and the spool displacement is shown. In order to ensure a comfortable ride, the spool displacement signal is fed back to the drive circuit 254 when the control input ∫εdt is zero to keep the spool displacement x _s at the neutral position (0%) to satisfy the comfortable ride. The orifice opening, that is, the damping coefficient C was given. The value of the damping coefficient C at that time is a function of the relative speed of the suspension.
Next, since the control input is also positive (+ ∫εdt) for the positive output (+ ε) of the multiplier 252, the spool displacement x _s is ε.
Oil passage 150 is fully closed from the neutral position according to (x _s = -100%)
Direction, the orifice opening a is reduced, the damping coefficient C is increased, and the damping force is increased. Further, for the negative output (−ε) of the multiplier 252, the control input is also negative (−∫εd
Therefore, the spool displacement x _s moves from the neutral position to the fully open ( _s = + 100%) direction of the oil passage 150 according to ε,
The orifice opening a is increased and the damping coefficient C is lowered to reduce the damping force.

 The operation of this embodiment is as follows.

In response to an external force or disturbance from the road surface, the microcomputer 270 detects the output v of the vehicle speed sensor 214, the wheel load w detected by the amplifier 221a, the relative displacement y detected by the linear potentiometer 210, and the temperature sensor 213. Relative displacement y, relative velocity, sprung displacement based on gas temperature t
x ₂ , sprung speed x ₂ , optimal gain G ₁ for damping force fc
G ₁₀ ) is output, and the optimum target control force u calculated based on the equation (35) is output from the adder 250. The deviation of the output of this target control force u from the damping force fc to be controlled is taken, and the difference is multiplied by the relative speed in the multiplier 252 to be changed into a damping force control signal, and the integrator is output according to the output. twenty five
3. By applying a current to the linear actuator 255 via the drive circuit 254 and moving the spool 258, the damping coefficient changes, and the damping force fc can be continuously changed.

The device of the embodiment of the present invention having the above-described operation can always select the optimum gain in the gas-liquid fluid suspension,
Since the damping force fc is continuously controlled based on the signal of the optimum target control force u calculated thereby, it is possible to adapt to any traveling state, and as a result, the riding comfort, traveling stability, etc. are improved significantly. There is an advantage that you can.

Further, in the multiplier 252 of the sign adjusting means II ₄ , the product of the deviation ε with respect to the target control force and the relative speed of the suspension ε
By doing so, the signal level rises compared to the deviation ε, so that the control signal ε having a good noise ratio to the signal, that is, the SN ratio can be obtained. In addition, the integrator 253 that integrates the signal in time eliminates the offset (residual deviation) that is harmful for controlling the fluffy component (0.2 Hz to 2 Hz) of sprung mass vibration that affects riding comfort to an optimum vibration level. You can Therefore, it is possible to control the damping force that follows the fluffy component of the target control force u, and to obtain an optimum vibration level, and at the same time, the gain is small for noise of high frequency that adversely affects the damping force control, and vibration control is possible. Since the gain is sufficiently high for the required frequency, there is an advantage that the stability of the control system can be improved.

Further, the actuator means IV for controlling the damping force fc is
Instead of using a return spring for the force generated by the linear actuator 255, the displacement of the spool 258 is fed back, so the spool 258 can be moved with a small amount of electrical energy, and the force generated thereby can be used effectively. As a result, the responsiveness is improved, fine vibrations with high frequency can be controlled, and power sources such as hydraulic pressure source and pneumatic pressure source are not required, and the weight, space and cost of piping etc. can be reduced. There is.

Although the sign adjusting means II ₄ of the present invention uses the multiplier 252, it may be a divider.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of the present invention, FIG. 2 is a block diagram showing a conventional technique, FIG. 3 is a diagram showing a four-degree-of-freedom model of a vibration system, and FIG. 4 is an embodiment of the present invention. Fig. 5 (a) is a diagram showing the position of the suspension in the vibration control device of the embodiment in which the present invention is applied to an automobile, and Fig. 5 (b) is a diagram showing the outline of the control mechanism of the embodiment, FIG. 6 is a block diagram showing an electronic control unit for controlling the control mechanism shown in FIG. 5 (b), FIG. 7 is an operation flow chart for explaining the operation of this embodiment, and FIG. 8 is FIG. For explaining the control of the single wheel portion in the embodiment of FIG. 9, FIG. 9 is an operation flow chart showing the operation flow of FIG. 7, and FIG. 10 (a) is an automobile gas-liquid fluid of the embodiment of the present invention. Fig. 11 (b) is a sectional view of the actuator means, and Fig. 11 is a control input. An orifice opening, the spool displacement characteristic diagram showing the relationship between the damping coefficient is.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Toshiyasu Minohe 1 Toyota-cho, Toyota-shi, Aichi Toyota Motor Co., Ltd. (56) References JP 62-108319 (JP, A) JP 62- 96114 (JP, A)

Claims (1)

(57) [Claims] 1. A state detecting means for detecting a physical quantity that influences a characteristic of a suspension supporting a vehicle, and a state quantity indicating a movement of the suspension and a state quantity indicating a movement of the vehicle, and the state detecting means. Detection force calculation means for calculating the detection control force corresponding to the physical quantity detected by, and a physical quantity representing the movement of the vehicle as a whole, which is the output of the state detection means, and an external force acting on the suspension, an external state such as a disturbance. And a predicted state calculation means for predicting and calculating the coming motion state of the entire vehicle based on a reference model corresponding to an average motion state during traveling corresponding to a motion state determined from the state quantity, and the predicted state calculation means. Based on the output of the vehicle, the maximum change in the motion state of the vehicle is accompanied by a predicted change in the state quantity of at least two or more states of the vehicle. A predictive state determining means for determining a dynamic state, and a vehicle control considering the detected control force in the reference model so as to control the variation of the motion state of the vehicle determined by the predictive state determining means to be minimum. Gain selection means for selecting the optimum feedback control gain calculated based on the model, and each gain selected by the gain selection means are multiplied by the physical quantity signal and the status quantity signal which are the outputs of the status detection means. And a control means including an optimum target control force calculation means for calculating an optimum target control force, and a deviation calculation means for calculating a deviation between the target control force and the detected control force, and the control means. Drive means for power-amplifying a deviation signal between the target control force and the detected control force, which is the output of the power supply, and acting on the suspension based on the power-amplified output. Actuator means for continuously variably controlling the characteristics of the suspension so as to equivalently generate a control force corresponding to a deviation of the detected control force actually detected with respect to the target control force in consideration of external force, disturbance, etc. From the degree of change of the state quantity or physical quantity of the suspension of the entire vehicle and each wheel, it is possible to appropriately determine the continuous change of the motion state of the vehicle, and to use the gain corresponding to the optimum target control. By calculating a force, an optimal target control force that is not affected by components such as a sudden external force and a disturbance with respect to the motion state of the vehicle is generated, and the characteristics of the suspension are continuously and optimally controlled. Variable damping force suspension control device.